In communication systems, the bit error rate (BER) is typically used as a figure of merit for an overall link function of the system. The BER is typically defined to be a measure of the number of errors that occur out of the total number of bits that have been transmitted over some finite amount of time. The BER is a statistical probability number, which is typically related to a signal-to-noise ratio (SNR) of the voltage, current, or light amplitude logic level of a transmission data signal. The BER is typically also related to time phase (jitter) amplitude of the data signal.
The BER may be determined using a BER tester (BERT), which typically includes a pattern generator and a pattern receiver (error detector), or the BER may be determined by analyzing the integrity of a data-eye diagram, which may also be referred to as a data-eye. The data-eye diagram is typically formed by overlapping data bits that occur over a finite period of time on a display, e.g., an oscilloscope. The BER performance can typically be correlated to the number of measured oscilloscope samples that do not comply with a data-eye mask (or an eye mask). The data-eye mask is typically user-specified to define boundaries or limits within which each data bit of the data-eye should fall in the absence of a bit error.
During production of communication systems, parametric performance values (e.g., output to input amplitude ratio vs. time, frequency, or wavelength) of the transmitter are typically optimized individually, and combined with one another to derive an optimum link BER. The BER optimization of each link component, as well as the overall link BER optimization, is typically achieved using feedback adjustment (based on previous measurements). The feedback adjustment typically involves adjusting certain circuit values, which may be electrical and/or optical, of active and/or passive circuit elements. After the feedback adjustment, the BER performance of each link component, as well as the overall link, is typically re-measured using external laboratory test instrumentation.
In a fiber-optic production test environment (a specific example of optical communication application), an optical transmitter is typically tested with a calibrated set of link components under simulated worst-case ambient, power supply, and medium conditions. The optical transmitter typically includes a phase locked loop (PLL) based data multiplexing, retiming, or repeating circuit, a laser diode driver (LDD), and a laser diode. The link components typically include fiber optic connector and cable, coupling optics, and an optical-to-electrical (O/E) converter. The link components typically are combined to form a looped data path connecting the calibrated pattern generator of a BERT system to the calibrated Error Detector of the same BERT system to test for the transmitter BER performance.
It is typically time-consuming and expensive to implement measurement and optimization functions for BER, data-eye, and any or all of the discrete parameters across all ambient, power supply, and medium, worst-case combinations. For example, it is often costly to build temperature control chambers and instrumentation that is designed to enable a human user to run the measurement and optimization procedures. Such systems often have much of their cost in the human-interface components (e.g., keypads, information display, and user interface software) In addition, it is often time-consuming to run man-in-the-loop tests over all worst-case combinations of ambient, power supply and medium characteristics. Further, it is often difficult to adjust the parameter values during operation in the field when changes to ambient and other conditions make such adjustments desirable.